The majority of gynecologic cancer incidence and mortality in the United States result from carcinoma of the uterine endometrium and epithelial ovarian carcinoma. These cancers arise from a multistep, multigenic process involving inherited and/or somatically acquired mutations. The long-term goals of this project are to define the most important molecular genetic alterations that contribute to the tumorigenic process for these two cancers. It is likely that elucidation of the spectrum of genetic mutations relevant to these cancers will contribute to their prevention and treatment through improved risk assessment, diagnostic, prognostic, and therapeutic strategies directed toward the critical genetic lesions. Germline mutations in a family of genes encoding DNA mismatch repair proteins have been found to cause the hereditary nonpolyposis colorectal cancer (HNPCC) syndrome. The tumors associated with this syndrome are characterized by genetic instability in simple sequence repeats (microsatellites) resulting from mutational inactivation of one of these DNA repair genes. Microsatellite instability (MIS) is also evident in a significant percentage of sporadic tumors that constitute the HNPCC syndrome, including cancers of the endometrium and ovary. The specific aims of this proposal are to determine the extent and characteristics of MIS in sporadic endometrial and ovarian carcinomas, and to address the hypothesis that the molecular basis for this phenomenon is the somatic mutational inactivation of one or more genes encoding DNA repair proteins, or alternatively, that MIS may also reflect an epigenetic phenomenon related to inappropriate cell division. To assess MIS, a large, representative panel of microsatellite markers will be examined for evidence of somatic length mutations in tumor compared to normal tissue DNA samples. For all tumors that exhibit MIS and for a matched set of tumors with no evidence of MIS, a mutational analysis will be carried out on the genes encoding DNA repair proteins that, when mutated, have been implicated in the induction of MIS. To investigate the potential role of cell proliferation as an epigenetic cause of MIS, estrogen-responsive human cells of known mismatch repair gene status will be treated with estrogens or mitogenic growth factors in vitro, and analyzed for the induction of MIS. Ancillary data relevant to these hypotheses will be generated through MIS quantitation and mismatch repair gene mutation analyses of endometrial hyperplasia specimens, a lesion that is believed to represent a precursor to type I, estrogen-related endometrial carcinomas. The same analyses will also be performed on cervicovaginal adenocarcinomas from women exposed in utero to the synthetic estrogen diethylstilbestrol. Together, these data should provide insight into the relative roles of somatic mismatch repair gene mutations vs. aberrant cell proliferation in the induction of MIS in gynecologic cancers.